This application is written in response to RFA DK-03-001, "Bench to Bedside Research in Type 1 Diabetes and Its Complications". The application represents a joint effort of two groups, one in Miami and one in Pittsburgh. The Miami group has a strong record of expertise and achievement in non-human primate and human islet transplantation, as well as fluency in islet transplant immunology. The Pittsburgh group has a strong track record using the transgenic and viral delivery of growth factors to the pancreatic islet. This application, therefore, in accord with the goals of the RFA, brings together basic scientists with expertise in beta cell molecular and cellular biology, beta cell gene therapy, and beta cell immunology, with animal physiologists expert in islet transgenic mouse development, rodent islet transplant, and non-human primate islet transplant, with clinicians with expertise in human islet transplantation and clinical management of diabetes. Specifically, the Pittsburgh group has shown that: 1) transgenic or adenoviral gene delivery of hepatocyte growth factor (HGF) to rodent islets prior to islet transplantation markedly improves transplant outcomes and islet graft function, 2) that this markedly reduces the number of islets required to achieve successful transplant outcomes in rodents, and 3) that HGF can be readily delivered to non-human primate islets using gene transfer techniques. The Miami group has demonstrated that the steroid-free immunosuppression protocol developed by the Edmonton group for humans can be successfully applied to non-human primates, and have developed an ideal marginal islet mass transplant model in non-human primates. This confluence of methodology and expertise allow the combined groups to test the hypothesis that delivery of HGF to non-human primate islets will very significantly enhance the engraftment and performance of these islets, and will therefore very significantly reduce the number of islets required to achieve normal glycemic control. We are optimistic that these studies will likely achieve their goal, and will serve in the near term as the third and final preclinical step required for similar studies in humans with diabetes.